A Young Exoplanet Caught at Formation Adam L. Kraus (Hubble Fellow; University of Hawaii-IfA) Michael Ireland (Macquarie University)

LkCa15 (Andrews et al. 2011)

LkHa 330 (Brown et al. 2009) The Case for Newly-Formed Planets Young planets are primordial Planets migrate based on interactions with the protoplanetary gas disk (left; Armitage & Rice 2005) or other planets and planetesimals.

Young planets are bright(er) Hot start or cold start aside, young planets should be orders of magnitude brighter (below; Marley et al. 2007; Baraffe et al. 2003). Sites of Ongoing Planet Formation

Transitional Disks Some protoplanetary disks have large cleared gaps – perhaps signposts of ongoing planet formation?

LkHa 330 (Submm image; Brown et al. 2009) Problem: These are very distant (~150-250 pc), so resolution is a big obstacle. CoKu Tau/4 (Model; Quillen et al. 2004) Nonredundant Mask Interferometry

Placing an aperture mask in the pupil plane turn the single mirror into a sparse interferometric array. Fourier analysis techniques (i.e. closure phases) can be used to filter almost all noise from turbulence and AO errors. This yields very stable performance. Closure Phases

Φ=Φ12+Φ23+Φ31

In the closure phase, any phase error on aperture 1, 2, or 3 will cancel out.

Readhead et al. (1988)

Almost all noise sources (most notably, speckles) are just phase errors that readily cancel. This improved PSF calibration delivers relatively deep contrast (~8 magnitudes in L’) even at the diffraction limit (70 mas at Keck). Following the Signposts of Planet Formation .

Andrews et al. (2011) Thalmann et al. 2010 SMA, Sub-mm Image HiCIAO, H-band Image (Dusty Outer Disk) LkCa 15 (Inner Disk Wall)

T~2-3 Myr; d~150 pc; M*=1 Msun; Mdisk=55 MJup LkCa15 b: A Planet in a Cleared Gap? Submm; Andrews et al. (2011) We see a central source that’s relatively blue (hot?) and leading/ trailing sources that are relatively red (cool?).

(The hot-start models say 6-12 MJup for the total resolved flux.)

11 AU (76 mas)

Nov 2009, L’ Nov 2010, L’ Nov 2010, K’ Multicolor Image of LkCa 15 b Submm; Andrews et al. in prep

(Deprojected. radius ~20 AU)

Andrews et al. (2011) Reconstructed Multicolor Image SMA, Sub-mm Image Blue=K’, Red=L’ Dusty Outer Disk Planet + Co-orbital material? Luminosity Evolution Models

Red: Disk Points show all free- Blue: No Disk floating Taurus members. Our detection falls at the very faint end (or below).

(K photometry from 2MASS, L photometry from Luhman et al. 2010.) Potential Explanations

 Reflected Light? LkCa 15 b  Geometry says MAYBE. Energetics and colors say NO. Total L’ flux is over 1% of the primary star flux!  Thermal disk emission  Much too far from the star, yet also too close to be the known inner wall. NO.  Planet (Deprojected radius ~20 AU)  Central flux is from a relatively blue planet (accretion luminosity?). The models would call it ~6 MJup, but be wary of accretion luminosity.  Leading/trailing flux is from co-orbital dust inside the Hill radius, perhaps accretion streams or tidal arms. This sounds promising. Implications and Future Directions

 Planet formation is not clean. Do not expect point sources. LkCa 15 b  Some planets form well outside the snow line. Disk instability? Does core accretion work at larger radii? Migration?  This planet is really bright! Hot-start model, or excess luminosity from accretion? (Deprojected radius ~20 AU)  When surveying transition disks, the planet detection rate is 1 out of 2. (And see Huelamo et al. 2011…)  Followup will be challenging. HKLM broadband colors with Keck-NRM, and GPI-NRM might give low-resolution spectra. Orbital motion in 3-5 years. Open questions

1.Submm; How Andrews did thiset al. in system prep form? Nominally too distant for core accretion, too close for disk instability.

2. Why is the gap large? (Rplanet~20 AU, Rgap~50 AU) 3. How is the extended material heated? Not direct radiation – shock heating from accretion, maybe?

(Deprojected. radius ~20 AU) GJ 802 B Observationally very similar: ΔK=4.9 ΔJ=5.2 ρ=50-100 mas

(Many points are actually from Palomar, with half the resolution!)

Visual and astrometric orbits are consistent; multi-epoch masking detections have been solid in every test we can find.

Blue: Masking data Red: Unresolved astrometry (Ireland et al. 2008) Companions in Fourier Space Single Mirror Amplitude Phase

F

Aperture Masked Kohler et al. (2000)

F + Closure Phases

Φ=Φ12+Φ23+Φ31

In the closure phase, any phase error on aperture 1, 2, or 3 will cancel out.

Readhead et al. (1988)

Almost all noise sources (most notably, speckles) are just phase errors that readily cancel. If you know the real phases to expect (i.e. from a 2 point source model), then you can fit the best model to fit the phase.